AI Briefing
KOSign in

Codync Launches as Open-Source Alternative to Grok Bot, Muse, and Dots

·2026.10.02 13:45

Key point

Codync is an MIT-licensed, Rust-based tool that lets users run 40+ coding agents like Claude Code and Codex as persistent bots, controlled remotely via iPhone, Mac, or Linux.

1 / 5

Details

Codync has launched as an open-source alternative to proprietary AI coding assistants like Grok Bot, Muse, and Dots. The tool allows developers to run 40+ coding agents (including Claude Code, Codex, and Cursor) as persistent, named bots on their local machines, which can then be managed and approved from mobile or desktop devices.

Key Features and Architecture

  • Open Source and Free: The project is MIT licensed, 100% free, and includes a free relay service for remote access. There are no subscriptions, paid tiers, or credit systems.
  • Agent Compatibility: Codync uses the Agent Client Protocol (ACP), enabling compatibility with a wide range of agents. Users can mix different agents in group chats or have one bot request reviews from another.
  • Native Performance: The host is a single Rust binary for Mac and Linux. Client apps are native: SwiftUI for iPhone and Mac, GTK for Linux, and a terminal UI over SSH. No web views are used.
  • Privacy and Security: All bots, transcripts, and memory remain on the user's local machine. Connections are end-to-end encrypted, either through the relay or directly via Wi-Fi/Tailscale, with notification text sealed to the phone's key.

Mobile and Desktop Integration

On the iPhone, users can approve edits and commands, group bots in rooms, and monitor activity via the Dynamic Island and Lock Screen. The app supports remote screen viewing and voice calls, with notifications triggered only when a bot requires approval or finishes a task. Installation on Mac is available via a one-line command or DMG, followed by pairing instructions for the iPhone.

This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.

Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.